Current efforts on AIDS vaccines are focused on 2 options: to make a vaccine that induces cellular immunity or to induce neutralizing antibodies. Up to now, all successful, licensed vaccines are based on antibodies, and many are not even capable of eliciting the cytolytic type of T cell immunity. However, at least 3 major obstacles have appeared to the antibody approach: virus variability, poor immunogenicity in humans,and relative insensitivity of fresh isolate to neutralizing antibodies. With regard to variability, we have demonstrated broadly reactive neutralizing antibodies in the sera of infected patients and have shown that these antibodies bind viral envelope gp120 and mapped the site to the conformation-dependent CD4 binding site. With regard to immunogenicity, we have demonstrated that-the problem is not due to gp120 interaction with human CD4, by showing that transgenic mice expressing human CD4 have the same response as normal mice lacking the transgene. However, during these studies, we identified host genes in the MHC region that do control the response of total antibodies, as well as antibodies to cross reactive determinants. High, intermediate, and low responders were identified, and we are working on expressing protein carrier vaccines to test the ability of carrier effects to overcome low responsiveness, as observed for other conjugate antigens. Finally, with regard to fresh isolates, we plan to study neutralizing antibodies against pseudovirions with macrophage-tropic envelopes. Preliminary studies indicate the feasibility of using pseudovirions to study infectivity of these viruses in cells expressing CKR-5, and we will now adapt this assay to measure neutralizing antibodies to these strains.